Post-Discharge Methicillin-Resistant Staphylococcus aureus Infections: Epidemiology and Potential Approaches to Control

From the Division of Adult Infectious Diseases, University of Colorado Denver, Aurora, CO, and the Department of Veterans Affairs, Eastern Colorado Healthcare System, Denver, CO.

Abstract

• Objective: To review the published literature on methicillin-resistant Staphylococcus aureus (MRSA) infections among patients recently discharged from hospital, with a focus on possible prevention measures.
• Methods: Literature review.
• Results: MRSA is a major cause of post-discharge infections. Risk factors for post-discharge MRSA include colonization, dependent ambulatory status, duration of hospitalization > 5 days, discharge to a long-term care facility, presence of a central venous catheter (CVC), presence of a non-CVC invasive device, a chronic wound in the post-discharge period, hemodialysis, systemic corticosteroids, and receiving anti-MRSA antimicrobial agents. Potential approaches to control include prevention of incident colonization during hospital stay, removal of nonessential CVCs and other devices, good wound debridement and care, and antimicrobial stewardship. Hand hygiene and environmental cleaning are horizontal measures that are also recommended. Decolonization may be useful in selected cases.
• Conclusion: Post-discharge MRSA infections are an important and underestimated source of morbidity and mortality. The future research agenda should include identification of post-discharge patients who are most likely to benefit from decolonization strategies, and testing those strategies.

Methicillin-resistant Staphylococcus aureus (MRSA) is a leading cause of morbidity and mortality due to infections of the bloodstream, lung, surgical sites, bone, and skin and soft tissues. The mortality associated with S. aureus bloodstream infections is 14% to 45% [1–4]. A bloodstream infection caused by MRSA is associated with a twofold increased mortality as compared to one caused by methicillin-sensitive S. aureus [5]. MRSA pneumonia carries a mortality of 8%, which increases to 39% when bacteremia is also present [6]. S. aureus bloodstream infection also carries a high risk of functional disability, with 65% of patients in a recent series requiring nursing home care in the recovery period [7]. In 2011 there were more than 11,000 deaths due to invasive MRSA infection in the United States [8]. Clearly S. aureus, and particularly MRSA, is a pathogen of major clinical significance.

Methicillin resistance was described in 1961, soon after methicillin became available in the 1950s. Prevalence of MRSA remained low until the 1980s, when it rapidly increased in health care settings. The predominant health care–associated strain in the United States is USA100, a member of clonal complex 5. Community-acquired MRSA infection has garnered much attention since it was recognized in 1996 [9]. The predominant community-associated strain has been USA300, a member of clonal complex 8 [10]. Following its emergence in the community, USA300 became a significant health care–associated pathogen as well [11]. The larger share of MRSA disease remains health care–associated [8]. The most recent data from the Center for Disease Control and Prevention Active Bacterial Core Surveillance system indicate that 77.6% of invasive MRSA infection is health care–associated, resulting in 9127 deaths in 2011 [8].

This article reviews the published literature on MRSA infections among patients recently discharged from hospital, with a focus on possible prevention measures.

MRSA Epidemiologic Categories

Epidemiologic investigations of MRSA categorize infections according to the presumed acquisition site, ie, in the community or in a health care setting. Older literature refers to nosocomial MRSA infection, which is now commonly referred to as hospital-onset health care–associated (HO-HCA) MRSA. A common definition of HO-HCA MRSA infection is an infection with the first positive culture on hospital day 4 or later [12]. Community-onset health care–associated MRSA (CO-HCA MRSA) is defined as infection that is diagnosed in the outpatient setting, or prior to day 4 of hospitalization, in a patient with recent health care exposure, eg, hospitalization within the past year, hemodialysis, surgery, or presence of a central venous catheter at time of presentation to the hospital [12]. Community-associated MRSA (CA-MSRSA) is infection in patients who do not meet criteria for either type of health care associated MRSA. Post-discharge MRSA infections would be included in the CO-HCA MRSA group.

Infection Control Programs

Classic infection control programs, developed in the 1960s, focused on infections that presented more than 48 to 72 hours after admission and prior to discharge from hospital. In that era, the average length of hospital stay was 1 week or more, and there was sufficient time for health care–associated infections to become clinically apparent. In recent years, length of stay has progressively shortened [13]. As hospital stays shortened, the risk that an infection caused by a health care–acquired pathogen would be identified after discharge grew. More recent studies have documented that the majority of HO-HCA infections become apparent after the index hospitalization [8,14].

Data from the Active Bacterial Core Surveillance System quantify the burden of CO-HCA MRSA disease at a national level [8,14]. However, it is not readily detected by many hospital infection surveillance programs. Avery et al studied a database constructed with California state mandated reports of MRSA infection and identified cases with MRSA present on admission. They then searched for a previous admission, within 30 days. If a prior admission was identified, the MRSA case was assigned to the hospital that had recently discharged the patient. Using this approach, they found that the incidence of health care–associated MRSA infection increased from 12.2 cases/10,000 admissions when traditional surveillance methods were used to 35.7/10,000 admissions using the revised method of assignment of health care exposure [15]. These data suggest that post-discharge MRSA disease is underappreciated by hospital infection control programs.

Lessons from Hospital-Onset MRSA

The morbidity and mortality associated with MRSA have led to the development of vigorous infection control programs to reduce the risk of health care–associated MRSA infection [16–18]. Vertical infection control strategies, ie, those focused on MRSA specifically, have included active screening for colonization, and nursing colonized patients in contact precautions. Since colonization is the antecedent to infection in most cases, prevention of transmission of MRSA from patient to patient should prevent most infections. There is ample evidence that colonized patients contaminate their immediate environment with MRSA, creating a reservoir of resistant pathogens that can be transmitted to other patients on the hands and clothing of health care workers [19,20]. Quasi-experimental studies of active screening and isolation strategies have shown decreases in MRSA transmission and infection following implementation [18]. The only randomized comparative trial of active screening and isolation versus usual care did not demonstrate benefit, possibly due to delays in lab confirmation of colonization status [21]. Horizontal infection control strategies are applied to all patients, regardless of colonization with resistant pathogens, in an attempt to decrease health care–associated infections with all pathogens. Examples of horizontal strategies are hand hygiene, environmental cleaning, and the prevention bundles for central line–associated bloodstream infection.

The Burden of Community-Onset MRSA

CO-HCA MRSA represents 60% of the burden of invasive MRSA infection [8]. While this category includes cases that have not been hospitalized, eg, patients on hemodialysis, post-discharge MRSA infection accounts for the majority of cases [15]. Recent data indicate that the incidence of HO-HCA MRSA decreased 54.2% between 2005 and 2011 [8]. This decrease in HO-HCA MRSA infection occurred concurrently with widespread implementation of vigorous horizontal infection control measures, such as bundled prevention strategies for central line–associated bloodstream infection and ventilator-associated pneumonia. The decline in CO-HCA MRSA infection has been much less steep, at 27.7%. The majority of the CO-HCA infections are in post-discharge patients. Furthermore, the incidence of CO-HCA MRSA infection may be underestimated [15].

Post-Discharge MRSA Colonization and Infection

Hospital-associated MRSA infection is reportable in many jurisdictions, but post-discharge MRSA infection is not a specific reportable condition, limiting the available surveillance data. Avery et al [15] studied ICD-9 code data for all hospitals in Orange County, California, and found that 23.5/10,000 hospital admissions were associated with a post-discharge MRSA infection. This nearly tripled the incidence of health care–associated MRSA infection, compared to surveillance that included only hospital-onset cases. Future research should refine these observations, as ICD-9 code data correlate imperfectly with chart reviews and have not yet been well validated for MRSA research.

The CDC estimated that in 2011 there were 48,353 CO-HCA MRSA infections resulting in 10,934 deaths. This estimate is derived from study of the Active Bacterial Core surveillance sample [8]. In that sample, 79% of CO-HCA MRSA infections occurred in patients hospitalized within the last year. Thus, we can estimate that there were 34,249 post-discharge MRSA infections resulting in 8638 deaths in the United States in 2011.

MRSA colonization is the antecedent to infection in the majority of cases [22]. Thus we can assess the health care burden of post-discharge MRSA by analyzing colonization as well as infection. Furthermore, the risk of MRSA colonization of household members can be addressed. Lucet et al evaluated hospital inpatients preparing for discharge to a home health care setting, and found that 12.7% of them were colonized with MRSA at the time of discharge, and 45% of them remained colonized for more than a year [23]. Patients who regained independence in activities of daily living were more likely to become free of MRSA colonization. The study provided no data on the risk of MRSA infection in the colonized patients. 19.1% of household contacts became colonized with MRSA, demonstrating that the burden of MRSA extends beyond the index patient. None of the colonized household contacts developed MRSA infection during the study period.